In this study,fresh pork tenderness,drip-loss,p H value and color parameters( CIE,a*,b*and L*values) were simultaneously predicted using hyperspectral scattering imaging( HSI) technique. The hyperspectral scattering images of 40 fresh pork samples were collected at the wavelength of 400- 1 100 nm,and the scattering profiles were fitted via Lorentzian distribution( LD) function to give three parameters a( asymptotic value),b( peak value) and c( full width at b /2). Stepwise discrimination was performed to determine the optimal wavelengths combinations. The LD parameters combinations( a,b and c) of optimal wavelengths were used to establish multi-linear regression( MLR) models to predict the pork attributes. The models were able to predict pork with high correlation coefficients of 0. 92 for drip-loss,0. 94,0. 92 and 0. 98 respectively for color parameters( a*,b*and L*),and for tenderness and p H value the models gave the correlation coefficients of 0. 69 and 0. 76,respectively. These results showed that the hyperspectral scattering technique was capable of predicting quality parameters of pork. The study provides an efficient means for rapid and nondestructive determination of pork quality simultaneously.
展开▼
